Electrostatic Assist of Liquid Transfer between Flat Surfaces

However, the contact line is still free to move when Eq. (26) is applied [23]. As a consequence no boundary condition needs to be applied on h. When t...
0 downloads 8 Views 5MB Size
Subscriber access provided by Oregon Health & Science University Library

New Concepts at the Interface: Novel Viewpoints and Interpretations, Theory and Computations

Electrostatic Assist of Liquid Transfer between Flat Surfaces Chung-Hsuan Huang, and Satish Kumar Langmuir, Just Accepted Manuscript • Publication Date (Web): 21 Mar 2018 Downloaded from http://pubs.acs.org on March 21, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Electrostatic Assist of Liquid Transfer between Flat Surfaces Chung-Hsuan Huang and Satish Kumar∗ Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN 55455, USA

Abstract - Transfer of liquid from one surface to another plays a vital role in printing processes. During liquid transfer, a liquid bridge is formed and subjected to substantial extension, but incomplete liquid transfer can produce defects that are detrimental to the operation of printed electronic devices. One strategy for minimizing these defects is to apply an electric field, a technique known as electrostatic assist (ESA). However, the physical mechanisms underlying ESA remain a mystery. To better understand these mechanisms, slender-jet models are developed for both perfect dielectric and leaky dielectric axisymmetric Newtonian liquid bridges with moving contact lines. Nonlinear partial differential equations describing the evolution of the bridge radius and interfacial charge are derived, and then solved using finite-element methods. For perfect dielectrics, application of an electric field enhances liquid transfer to the more wettable surface over a wide range of capillary numbers. The electric field modifies the pressure differences inside the liquid bridge, and as a consequence, drives liquid toward the more wettable surface. For leaky dielectrics, charge can accumulate at the liquid-air interface. Application of an electric field can augment or oppose the influence of wettability differences, depending on the direction of the electric field and the sign of the surface charge. Flow visualization experiments reveal that when an electric field is applied, more liquid is transferred to the more wettable surface due to a modified bridge shape that causes depinning of the contact line. The measured values of the amount of liquid transferred are in good agreement with predictions of the perfect dielectric model.



Electronic mail: [email protected]

1 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Introduction

Although printing processes have been used for centuries to make books and newspapers, they are increasingly being developed for large-scale manufacturing of electronic devices including transistors [1, 2, 3, 4], solar cells [5], and antennas [6]. In these printing processes, the inks are conductive materials that are deposited onto flexible substrates such as paper [7] or thin plastic films [8]. A flexible substrate is unwound from a cylindrical roll and then wound onto another roll after printing. Such roll-toroll printing techniques have potential to significantly reduce production costs, speed up manufacturing processes [9], and reduce waste disposal [10]. One important printing process is gravure printing (Fig. 1(a)). The gravure roll is a metal cylinder that has been engraved with a large number of micron-scale cavities. The cavities are filled with ink from a liquid pool, and the ink then transfers from the cavities to a substrate. A backing roll is used to force contact between the substrate and gravure roll. For printed electronics, it is essential that enough ink be transferred from the cavities to the substrate to prevent broken circuits and failure of electronic devices. However, it has been found that only 50% of the ink transfers to the substrate at high printing speeds. An electrostatic potential difference between the substrate and gravure roll is sometimes applied to improve ink transfer in traditional printing processes [11, 12, 13, 14], a method is known as electrostatic assist (ESA) (Fig. 1(a)). However, the physical mechanisms underlying ESA remain a mystery. If these mechanisms were better understood, then ESA could be effectively applied to the printing of electronic devices, thereby reducing the number of defects. This paper describes research addressing this issue. The key to printing processes involves the formation of a liquid bridge, which is a mass of liquid connecting two solid surfaces, and transfer of liquid from one surface to another (Fig. 1(b)). This paper focuses on an important model problem concerning liquid transfer, shown in Fig. 1(c). A mass of liquid is initially deposited on the bottom surface. A second flat surface is lowered so that it comes into contact with 2 ACS Paragon Plus Environment

Page 2 of 42

Page 3 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 1: (a) Schematic of a gravure printing process with electrostatic assist. The ink in a cavity is transferred from an engraved roll to a substrate. (b) Illustration of a stretching liquid bridge. (c) Schematic of a stretching liquid bridge subject to an electric field. that liquid to form a liquid bridge. An electrostatic potential difference is applied between the two flat surfaces, and the top surface is then pulled vertically until the liquid bridge breaks. During this process, the contact lines (where air, liquid, and solid meet) move on the two flat surfaces. The central issue is to determine the amount of liquid transferred from the bottom surface (printing plate) to the top surface (substrate) as the two surfaces separate. Although gravure printing involves liquid transfer from cavities (Fig. 1(b)), the model problem considered here (Fig. 1(c)) provides a necessary foundation for understanding the more complex problem involving cavities. This model problem is also highly relevant to other applications including adhesion [15], oil recovery [16], and microfluidics [17]. There have been a number of prior studies investigating liquid transfer between two flat surfaces subject to extensional motion in the absence of an electric field [9, 18, 19, 20, 21, 22]. A key parameter characterizing the problem is the capillary number, Ca = µU/γ, which represents the ratio of viscous forces to surface-tension forces, where µ is the liquid viscosity, U is the stretching speed, and γ is the liquid-air surface tension. Experimental results and model predictions show that when Ca is smaller than ∼10−2 , liquid transfer is determined by the wettability difference between the two surfaces. When Ca is ∼1, viscous forces become stronger and reduce the influence of wettability differences. As a consequence, the amount of liquid transferred is close to 50%. It is thus a major challenge to transfer more than 50% of the liquid at 3 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

high printing speeds. Other forces, such as those arising from applied electric fields (Fig. 1(a)), could be an effective way to enhance liquid transfer at high capillary numbers. Because the liquid bridge becomes long and thin during stretching, slender-jet models have been applied to describe bridge dynamics [22, 23, 24, 25]. These models are based on the assumption that the variation of the interface shape is small relative to the bridge length, so to leading order the bridge shape and axial velocity depend only on the axial coordinate and time [24]. Predictions from slender-jet models of the amount of liquid transferred agree qualitatively and sometimes quantitatively with experimental observations [22, 25]. Slender-jet models thus provide an efficient way to describe the shape of stretching liquid bridges with moving contact lines. Electrostatic effects have been incorporated into slender-jet models to describe electrospinning [26, 27, 28], and the goal of the present paper is to make a similar effort to describe liquid transfer. The paper is organized as follows. The mathematical model, which consists of an axisymmetric Newtonian liquid bridge between two surfaces of different wettability that are subject to an electric field and can separate vertically (Fig. 1(c)), is developed in Sec. 2. Results for perfect dielectric (non-conducting) liquids and leaky dielectric (weakly conducting) liquids are presented in Secs. 3 and 4, respectively. Complementary flow visualization experiments have also been performed, and these are reported and compared to the model predictions in Sec. 5. In Sec. 6, we summarize our results and provide some concluding remarks.

2

Problem formulation

An axisymmetric Newtonian liquid bridge with constant density ρ, viscosity µ, and surface tension γ is confined between two flat electrodes (Fig. 1(c)). Air is outside the bridge and assumed to have no influence on the liquid flow. The top electrode has a constant electrostatic potential φ∗ and is moved in the z-direction (vertically) with a constant velocity U , while the bottom electrode is grounded and remains stationary. 4 ACS Paragon Plus Environment

Page 4 of 42

Page 5 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Note that because the potentials are constant, the magnitude of the electric field decreases during bridge stretching. The contact angle for the top electrode is denoted as θtop , and that for the lower electrode is θbottom . Even though these contact angles may depend on the speed of contact line, we consider here the limiting case where they are constant. In this study, we consider two types of liquids. The first is a perfect dielectric, for which the conductivity and charge are zero but polarization can occur. The other is a leaky dielectric [29, 30, 31], which is polarizable but weakly conductive. Charges inside the liquid accumulate at the liquid-air interface and are assumed to be negligible in the bulk. Air is modeled as a perfect dielectric fluid with relative permittivity  = 1. The relative permittivity of the liquid is  = ∗ , its conductivity is K, and the free-space permittivity is o .

2.1

Governing equations

We scale length by a characteristic initial bridge radius R, velocity by U , time by R/U , pressure by µU/R, electrostatic potential by φ∗ , electric field by E ∗ = φ∗ /L(0) (where L(0) is the initial bridge length), charge density by o E ∗ , and conductivity by K ∗ = RK/U o . The mass conservation equation is ∇ · v = 0.

(1)

Electric charge conservation at the liquid-air interface requires that ∂σ + v · ∇s σ = σn · (n · ∇)v+ k −K ∗ E k ·n, ∂t

(2)

where σ is the surface charge density, t is time, ∇s = (I − nn) · ∇ is the surface gradient operator, E is the electric field, k ... k is the jump operator which yields the quantities in the air phase minus those in liquid phase, and n is the outward unit normal to the interface. The electric field is related to the electrostatic potential through E = −∇φ. 5 ACS Paragon Plus Environment

(3)

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 42

The momentum conservation equation is Re(

∂v + v · ∇v) = −∇P + χ∇ · σ M + ∇2 v, ∂t

(4)

where v is the velocity vector and P is the pressure. The two parameters that appear are the electroviscous number χ = RE ∗2 o /µU , which provides a ratio of electrostatic to viscous forces, and the Reynolds number Re = ρU R/µ, which provides a ratio of inertial to viscous forces. In Eq. (4), σ M is the Maxwell stress tensor which is given by o EE − 12 o E · EI, where I is the identity tensor. The divergence of the Maxwell stress tensor is zero for perfect and leaky dielectrics [31]. Hydrodynamic and electrostatic phenomena are coupled through the normal and tangential stress balances at the liquid-air interface [32]. The dynamics of the air phase are assumed negligible. At the liquid-air interface, we have the kinematic boundary condition, n · v = 0.

(5)

The normal stress balance is k n · T · n k= −

κ , Ca

(6)

where T = −P I + µ[∇v + (∇v)T ] + σ M is total stress tensor, κ = ∇s · n is the interfacial curvature, and Ca = µU/γ is the capillary number, which provides a ratio of viscous to surface-tension forces. The tangential stress balance is k n · T · t k= 0,

(7)

where t is the unit tangent vector to the liquid-air interface. Gravitational forces are assumed to be negligible relative to surface-tension forces, which corresponds to a small Bond number, Bo = ρgR2 /γ. At each solid surface, the no-slip and no-penetration conditions are twall · v = 0 at z = 0, L(t), 6 ACS Paragon Plus Environment

(8)

Page 7 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

( nwall · v =

0 for z = 0 1 for z = L(t),

(9)

where nwall is the normal vector pointing in the positive axial direction, twall is the tangential vector pointing in the radial direction, and L(t) represents the location of the top electrode. We note that near moving contact lines, the no-slip boundary condition breaks down and is usually replaced by a slip law, a point we address in Sec. 2.3. The electrostatic potential φ in each phase is governed by Laplace’s equation, ∇2 φ = 0. The boundary conditions for the top and bottom electrode are ( 0 for z = 0 φ= 1 for z = L(t).

(10)

(11)

The tangential components of the electric field are continuous at the interface, and a jump in the normal component of the electric field is related to the surface charge density σ [31], k E k ·n = σ.

2.2

(12)

1D slender-jet model

The liquid bridge radius is assumed to be much smaller compared to its length, so a small parameter s ≡ R/L(0) can be defined. We then perform an asymptotic expansion on the axial velocity v , axial electric field E , and pressure P [22, 23, 24]: v (r, z, t) = vo (z, t) + (s r )v1 (z, t) + (s r )2 v2 (z, t) + · · · ,

(13)

E (r, z, t) = Eo (z, t) + (s r )E1 (z, t) + (s r )2 E2 (z, t) + · · · ,

(14)

P (r, z, t) = Po (z, t) + (s r )P1 (z, t) + (s r )2 P2 (z, t) + · · · .

(15)

Because the dependence on the radial coordinate r is given explicitly in the expansion, the correction functions of these three variables only depend on time and the axial coordinate. 7 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 42

Equations (13)-(15) are substituted into Eqs. (1)-(12) to reduce them to a onedimensional (1D) problem [33]. After keeping the leading order terms and dropping the subscripts “o” for simplicity, the governing equations become Re(vt + vvz ) = −

1 3 2σE κz + 2 (h 2 vz )z + χ(βEEz + σσz + ), Ca h h

1 ht = −vhz − vz h, 2 β E − E∞ + ln(Λ)((σh)z − (Eh2 )zz ) = 0. 2 σ h σt + v σz + vz − K ∗ (hz E + Ez ) = 0, 2 2

(16) (17) (18) (19)

where h is the position of the liquid-air interface (i.e., bridge radius), E∞ is the electric field in the absence of the liquid, Λ = L(0)/2R is the initial bridge aspect ratio, β = ∗ − 1 is the permittivity constant, and κ is the interfacial curvature, κ=

1 h(1 +

1

hz2 ) 2



hzz 3

(1 + hz2 ) 2

.

(20)

The indices t and z indicate differentiation with respect to those variables. The full curvature is kept to gain a more accurate description of the interface [23, 24, 33, 34, 35, 36]. The electric field inside the bridge is determined by the induced charges or free charges at the interface. Equation (18), which has been previously used by Feng [26] and Hohman et al. [27] to describe electrospinning, is derived from an integral solution (in cylindrical coordinates) to Laplace’s equation [37].

2.3

Solution method

Equations (16)-(19) are solved with a finite-element method [22, 23, 24]. To lower the highest-order derivative in these equations, two new variables, Ω and ω, are introduced, ∂h = 0, ∂z ∂(h2 E) = 0. ω− ∂z Ω−

8 ACS Paragon Plus Environment

(21) (22)

Page 9 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

The no-penetration boundary condition is v (z = 0, t) = 0, v (z = L(t), t) = 1.

(23)

The boundary conditions on electrostatic potential are φ(z = 0, t) = 0, φ(z = L(t), t) = 1.

(24)

No-flux boundary conditions are applied on σ, ∂σ ∂σ (z = 0) = 0, (z = L(t)) = 0. ∂z ∂z

(25)

The contact angle appears through a geometric relation Ω=

∂h = cot(θ), ∂z

(26)

where θ is θtop or θbottom . It turns out that the slip law (§2.1) does not appear at leading order and so cannot be imposed [23]. However, the contact line is still free to move when Eq. (26) is applied [23]. As a consequence no boundary condition needs to be applied on h. When the liquid bridge is about to pinch off, the droplet on each surface is connected by a thin liquid thread. We stop the simulations when the narrowest bridge radius is 1% of its initial radius. At this point, the overall bridge shape has largely been set, and thus the transfer ratio, which is the percentage of liquid transferred to the top surface (plate), does not change significantly even if the bridge thins further. To ensure mass conservation, we monitor the variation of volume, Vvar = 100(Vmax −Vmin )/Vini , where Vmax , Vmin , and Vini are the maximum, minimum and initial bridge volume, respectively. If Vvar is larger than 0.1%, we increase the number of elements or decrease the time step, then rerun the simulation and repeat this process until Vvar is less than 0.1% [21].

9 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2.4

Page 10 of 42

Parameter values

Order-of-magnitude values of the dimensional parameters are listed in Table 1, and the corresponding dimensionless parameters are shown in Table 2. In this work we choose to focus on regime of small Bond numbers (Bo) and thus neglect gravitational effects, as we have done in Eqs. (4) and (16). Similarly, we choose to focus on the regime of small Reynolds numbers (Re), and thus neglect inertial forces by setting Re = 0. We thus investigate the influence of the capillary number, Ca, and electroviscous number, χ. The dimensionless conductivity, K ∗ , is the ratio of the time scale for bridge stretching to the charge relaxation time and appears when considering leaky dielectrics. In our calculations, the initial bridge aspect ratio (ratio of initial length to initial diameter) is around unity. As the bridge stretches, the aspect ratio increases and the slender-jet model is expected to provide an increasingly accurate approximation. Table 1: Order-of-magnitude values of dimensional parameters

3 3.1

Parameter

Typical Values

Bridge radius R and bridge length L (mm)

10−3 − 1

Liquid viscosity µ (cP)

1 − 102

Surface tension γ (mN/m)

101 − 102

Stretching speed U (mm/s)

1 − 102

Voltage V (volt)

102 − 104

Conductivity K (S/m)

10−12 − 10−2

Results: Perfect dielectrics Liquid transfer

In the absence of an electric field, the transfer ratio (percentage of liquid transferred to the top plate) is determined by the wettability difference between the two plates when 10 ACS Paragon Plus Environment

Page 11 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Table 2: Dimensionless problem parameters Parameter

Definition

Physical meaning

Value

Bo

ρgR2 /γ

Gravitational forces Surface-tension forces

10−7 − 1

Re

ρU R/µ

Inertial forces Viscous forces

10−5 − 102

Ca

µU/γ

Viscous forces Surface-tension forces

10−5 − 1

χ

REo2 o /µU

Electrostatic forces Viscous forces

10−2 − 108

Relative permittivity

1 − 101

 β

Polarizability difference between

∗ − 1

1 − 101

liquid and air phases K∗

(R/U )/(o /K)

Time scale for stretching Time scale for charge relaxation

10−6 − 109

Λ

L(0)/2R

Initial bridge length Initial bridge diameter

1

Ca is low (< 0.01) [21, 22]. As Ca increases, viscous forces become stronger relative to surface-tension forces, and as a result, the transfer ratio approaches 50% regardless of the wettability difference. Here, we explore how an electric field influences this behavior for perfect dielectric liquids. This corresponds to setting σ = 0 and K = 0, and solving Eqs. (16)-(18). In this section, the initial shape is taken to be a cylinder with length L(0) and radius R. This choice allows us to isolate the influence of the electric field; the influence of initial bridge shape is explored in Sec. 3.3. The initial cylinder aspect ratio (Λ = L(0)/2R) is 1. The capillary number Ca is set to 0.1 or 1, and the electroviscous number χ ranges from 0 (no electric field applied) to 100. The permittivity constant β = 1.74, which corresponds to the permittivity of silicone oil. We choose this value because silicone oil has been used in experiments examining the stability of liquid bridges in electric fields [38, 39, 40, 41]. For computational convenience, we rotate the liquid bridge 90◦ clockwise, so the top/bottom plate is on the right/left side in the figures showing bridge shapes. Figure 2 shows the transfer ratio for the cases where the top contact angle θtop = 60◦ and the bottom contact angle θbottom ranges from 40◦ to 90◦ . These angles are 11 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2: Relationship between transfer ratio and wettability difference for different values of χ for (a) Ca = 0.1 and (b) Ca = 1. Here, θtop is 60◦ and θbottom varies from 40◦ to 80◦ . The wettability difference is ∆θr = θbottom − θtop , and the parameter β = 1.74. assumed to be receding contact angles [22]. The wettability difference ∆θr = θbottom − θtop , so a positive ∆θr means the top plate is more wettable than the bottom plate. When Ca = 0.1, the transfer ratio increases as ∆θr increases when χ = 0 (Fig. 2(a)). Because the bottom plate becomes less wettable when ∆θr increases, more liquid is transferred to the top plate. When Ca is O(1) and no electric field is present (Fig. 2(b)), viscous forces are stronger and reduce the effect of surface wettability differences, so the transfer ratios are nearly 50% for all values of ∆θr . We note that for values of Ca < 0.1, the influence of the electric field is considerably weaker since χCa, the ratio of electrostatic forces to surface tension forces, becomes smaller. Both Figs. 2(a) and 2(b) show that when χ increases from 0 to 50, the transfer ratio of the electrified bridge is higher than that for the non-electrified bridge when ∆θr is positive. However, when ∆θr is negative, the presence of an electric field decreases the amount of liquid transferred. These results show that the electrostatic forces enhance liquid transfer to the more wettable surface even when viscous forces are strong (Ca = O(1)). In Fig. 2(a), when χ is doubled from 50 to 100, the transfer ratio only slightly increases or decreases. Therefore, we will focus on the cases χ = 0 and 50.

12 ACS Paragon Plus Environment

Page 12 of 42

Page 13 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

As mentioned in the introduction, the transfer ratio is only about 50% at high printing speeds, and this represents a major challenge to print quality. Figure 2(b) shows that application of an electric field can significantly increase the transfer ratio when ∆θr is positive and Ca is O(1). This finding is consistent with the use of ESA (Sec. 1) to improve liquid transfer in industrial printing processes.

Figure 3: Comparison of final bridge shapes and contact-line positions for the cases χ = 0 (solid lines) and 50 (dashed lines). (a,c) Ca = 0.1; the breakup times for χ = 0 and 50 are 2.1 and 2.4, respectively. (b,d) Ca = 1; the breakup times for χ = 0 and 50 are 8.8 and 7.6, respectively. Values of other parameters are θbottom = 70◦ , θtop = 60◦ , ∆θr = 10◦ , and β = 1.74. We now characterize how the electric field changes the bridge shape and contactline movement. Figures 3(a) and 3(b) shows the final bridge shapes when χ = 0 (no electric field) and 50 for Ca = 0.1 and 1. The wettability difference ∆θr is 10◦ . In both Figs. 3(a) and 3(b), the contact line positions on the top (right) plate for both χ = 0 and 50 are almost the same. However, the contact line of the electrified 13 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

liquid bridge on the bottom (left) plate (which is less wettable than the top plate) has slipped more relative to that of the non-electrified liquid bridge. This result shows that the presence of the electric field increases the transfer ratio by pushing more liquid away from the less wettable surface. The mechanism for this will be discussed in Sec. 3.2. Figures 3(c) and 3(d) show the time-evolution of the contact-line positions. In Fig. 3(c) (Ca = 0.1), both the top and bottom contact radii (i.e., radii at top/bottom plates) of the electrified bridge are always smaller than those of the non-electrified bridge, but the top contact radius of the electrified bridge eventually reaches a similar value to that of the non-electrified bridge. However, the bottom contact line of the electrified bridge retreats much faster than that of the non-electrified bridge. Therefore, less liquid remains on the bottom plate (Fig. 3(a)) and the transfer ratio increases when the electric field is present. Similar behavior is seen in Fig. 3(d) (Ca = 1), although here the top contact radius of the electrified bridge can become larger than that of the non-electrified bridge. The contact radii increase with time at later times because of the high pressure that develops near the bridge center. The resulting pressure gradients drive liquid toward the contact lines, causing the contact radii to increase. In practice, the contact lines could pin during bridge stretching due to surface heterogeneities. We briefly discuss the influence of the electric field when one contact line is pinned. When both contact lines are pinned, the electric field does not affect the transfer ratio due to symmetry. Figure S1 shows transfer ratio with different bottom contact angles when the top contact line is pinned and Ca = 1. We only discuss the case where Ca = 1 because the effect of the electric field on liquid transfer at lower Ca is qualitatively similar. When χ = 0, the transfer ratio is around 50%. This can be understood by recognizing that when viscous forces are strong enough, the contact line does not slip much along the plate. The pinch-off point is near the center of the bridge, and the bottom contact line is close to its original position.

14 ACS Paragon Plus Environment

Page 14 of 42

Page 15 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

When χ = 50, the transfer ratio increases as the bottom contact angle increases. The reason is similar to the cases where both contact lines can move. The top plate can be rationalized as a surface with such a small receding contact angle that the contact angle on the top plate does not reach the receding value and the contact line is pinned during stretching. The bottom contact line of the electrified bridge retreats much more than that of the non-electrified bridge. As a result, the transfer ratio increases when the electric field is present. However, if we pin the bottom contact line and allow the top contact line to slip, the results (not shown) are the mirror images of the case where the top contact line is pinned. Therefore, the transfer ratios are less compared to the case where there is no electric field and the bottom contact line is pinned. Figures 3(c), 3(d), and S1 illustrate that the presence of an electric field has a significant influence on contact-line motion, especially on the less wettable surface. As a consequence, the transfer ratio is significantly affected. We now discuss the physical mechanism underlying these phenomena by analyzing the pressure in the liquid bridge.

3.2

Physical mechanisms

For perfect dielectric liquids, the only way the Maxwell stress tensor affects the liquid bridge is through the normal stress balance (Eq. (6)). The resulting expression for the liquid pressure in the slender-jet approximation is P =

κ β − vz − χ E 2 . Ca 2

(27)

Note that in the slender-jet approximation, the pressure depends spatially only on the axial coordinate. The three terms on the right-hand side of Eq. (27) represent contributions from surface-tension, viscous, and electrostatic forces, respectively. The electrostatic term arises from the polarizability difference (β) between the liquid and air phase, and always acts to decrease the value of the pressure. Note that increasing the value of β will tend to enhance electrostatic effects. 15 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4: (a) Comparison of bridge shapes for the cases χ = 0 (solid line) and 50 (dashed line) at t = 1. (b) Comparison of relative pressure for the cases χ = 0 (solid line) and 50 (dashed line) at t = 1. Values of other parameters are Ca = 1, θbottom = 60◦ , θtop = 60◦ , ∆θr = 0◦ , and β = 1.74. In determining physical mechanisms, it is useful to define a relative pressure, which is the difference between the pressure and the minimum pressure in the liquid bridge. We set Ca = 1 because of its relevance to high-speed printing, and first consider the case ∆θr = 0◦ to remove the influence of wettability differences. Figure 4(a) compares bridge shapes when χ = 0 and 50 at t = 1, and the corresponding relative pressure profiles are shown in Fig. 4(b). In Fig. 4(a), the contact radii of the electrified bridge are smaller than those of the non-electrified bridge. In addition, the narrowest radius of the electrified bridge is greater than that of the nonelectrified bridge. These features can be explained by examining the pressure in the bridge (Fig. 4(b)). The pressure of both bridges is highest at the center and lowest at each contact line due to the axial variation of the radius. However, the highest pressure of the electrified bridge is smaller than that of the non-electrified bridge due to the presence of the electric field. In Eq. (27), the electrostatic term is always negative since χ and β are always positive. Note that the E in Eq. (27) represents the electric field tangent to the interface. The normal component of the electric field is negligible due to the slenderjet assumption. Therefore, the magnitude of the electrostatic term is greater near the center than near each contact line. As a consequence, the maximum pressure difference of the electrified bridge is smaller than that of the non-electrified bridge. 16 ACS Paragon Plus Environment

Page 16 of 42

Page 17 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

As the bridge extends, the contact radii decrease to conserve mass. However, since the bridge radius is narrowest at the center, liquid is pumped from the center toward the contact lines. When the electric field is applied, the smaller maximum pressure difference results in less liquid being driven away from the bridge center, leading to a larger value of the narrowest radius and smaller values of the contact radii.

Figure 5: Comparison of bridge shapes (a,c) for the cases χ = 0 (solid line) and 50 (dashed line) with corresponding relative pressures (b,d) at (a,b) t = 1 (c,d) t = 4. Values of other parameters are Ca = 1, θbottom = 70◦ , θtop = 60◦ , ∆θr = 10◦ , and β = 1.74. We now consider the case where ∆θr = 10◦ to examine how the electric field enhances liquid transfer to the more wettable surface. Figure 5(a) shows bridge shapes and the corresponding relative pressure profiles at t = 1. In Fig. 5(a), the bottom contact radius of the electrified bridge is smaller than that of the non-electrified bridge. In addition, the narrowest radius of the electrified bridge is located closer to the bottom (left) plate, which means that the bridge would break closer to the bottom plate when an electric field is applied. 17 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

As can be seen in Fig. 5(b), when the electric field is present, the difference between the maximum pressure and the pressure at the top (right) plate is larger than the difference between the maximum pressure and the pressure at the bottom (left) plate. Since the bottom contact angle is greater than the top contact angle, the magnitude of the tangential component of the electric field at the bottom is greater than that at the top. Therefore, the electric field has a greater influence on the pressure near the bottom plate, and the difference between the maximum pressure and the pressure at the bottom plate decreases significantly when an electric field is applied. As a consequence, liquid is driven from the region of maximum pressure (near the bottom plate) to the top plate, causing the bottom contact radius to decrease even more. The decrease in the bottom contact radius at later times (t = 4) is seen in Fig. 5(c). Because the bottom contact radius of the electrified bridge is considerably smaller than that of the non-electrified bridge, the electrified bridge has a smaller minimum bridge radius. As a result, the maximum relative pressure in the electrified bridge becomes larger than that in the non-electrified bridge, and so does the difference in relative pressures between the maximum and the top (right) plate, as seen in Fig. 5(d). The subsequent flow causes the electrified bridge to break earlier than the non-electrified bridge when Ca = 1 (Fig. 3(b)), and leads to more liquid being transferred to the more wettable surface. Note that since the magnitude of the electric field decreases during stretching (due to a constant voltage difference between the top and bottom plates), the electrostatic contribution to the pressure is not able to stabilize the bridge as breakup is approached. When Ca = 0.1, the mechanism for the increase in liquid transfer to the more wettable surface in the presence of the electric field is the same as in the case where Ca = 1 (results not shown). Because viscous forces are weaker relative to surfacetension forces (which drive liquid away from the narrowest part of the bridge) when Ca = 0.1, the bridge cannot be stretched as far before breakup compared to the case where Ca = 1 (Fig. 3). However, the shorter breakup length means that electrostatic forces are more effective at stabilizing the bridge. As a result, the electrified bridge 18 ACS Paragon Plus Environment

Page 18 of 42

Page 19 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

can be stretched longer than the non-electrified liquid bridge when Ca = 0.1 (or lower) (Fig. 3(a)). When the two surfaces have the same wettability (∆θr = 0◦ ), we also observe that an electrified bridge cannot be stretched as far before breakup compared to a nonelectrified bridge when Ca = 1 (results not shown). At early times, the maximum pressure in the electrified bridge (which occurs at the point of narrowest radius) is lower compared to that in the non-electrified bridge due to the stabilizing effect of the electric field. However, at later times, the maximum pressure in the electrified bridge becomes larger than that in the non-electrified bridge due to the weakening of electrostatic forces as the bridge length increases. The larger maximum pressure leads to larger pressure gradients that rapidly drive liquid away from the narrowest part of the bridge, leading to a shorter breakup time for the electrified bridge.

3.3

Influence of initial bridge shape

In general, the initial liquid bridge shape in experiments will not be a perfect cylinder [18, 42]. For example, if an equilibrium shape is used, the contact line positions on each surface will be different due to differences in surface wettability. In this section, we report results from a brief parametric study examining the influence of using equilibrium bridge shapes as initial conditions. To obtain an equilibrium bridge shape, we start with a cylinder having a length equal to its diameter (Λ = 1), set the velocity of both plates to zero, and solve the governing equations numerically until an equilibrium state is reached. Because the velocity of the top plate is zero, a new velocity scale is needed. We choose this to be γ/µ, and the corresponding time scale is Rµ/γ. Therefore, a new dimensionless parameter representing the ratio of electrostatic forces to surface-tension forces arises, the electrocapillary number CaE = REo2 o /γ. Equation (16) now becomes 0 = −κz +

3 2 (h vz )z + CaE βEEz , h2

but Eqs. (17) and (18) remain the same. 19 ACS Paragon Plus Environment

(28)

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 42

Figure S2(a) shows equilibrium bridge shapes for CaE = 0, 10, and 50 when ∆θr is 0◦ . The results are similar to those in Fig. 4. The electrostatic contribution to the pressure reduces the pressure difference in the bridge, leading to flatter shape as CaE increases. (The initial radius is 1). However, if we use these equilibrium shapes as initial conditions and stretch the bridge, the transfer ratio is still 50% due to the absence of a wettability difference. Therefore, we use the equilibrium bridge shapes shown in Fig. S2(b), where ∆θr = 10◦ , as initial conditions. We again focus on relatively high stretching speeds and set Ca = 1. In Fig. S2(b), because the top surface is more wettable, the top contact radius is larger than the bottom contact radius when both the electrified and non-electrified bridges reach equilibrium. As a consequence, the narrowest bridge radius is closer to the bottom plate in both cases. When the equilibrium shape is taken to be the initial condition, the transfer ratio of the non-electrified bridge is 72%, and the transfer ratio of the electrified bridge is 79% (Table 3). The transfer ratio of both bridges increases compared to the case where perfect-cylinder initial conditions are used. Even though the transfer ratio of the non-electrified bridge increases about 20% with this change in initial condition, the transfer ratio of the electrified bridge is still larger than that of the non-electrified bridge. Table 3: Comparison of transfer ratios for different initial bridge shapes Initial shape

χ=0

χ = 50

Perfect cylinder

53%

76%

Equilibrium shape

72%

79%

The transfer ratio of the non-electrified bridge increases about 20% because more than 50% of the liquid is initially close to the top surface. Figure S2(b) shows the top contact radius of non-electrified liquid bridge is 30% larger than its bottom contact radius. Since the contact-line motion is relatively small due to the weaker surfacetension forces when Ca = 1, the initial shape has a significant influence on the final bridge shape and thus the transfer ratio.

20 ACS Paragon Plus Environment

Page 21 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

The transfer ratio of the electrified bridge only increases 3% when an equilibrium initial shape is used. Because the electrostatic contribution to the pressure leads to a flatter bridge shape, the contact radii are still close to 1 (Fig. S2(b)) like with a perfect-cylinder initial shape. However, when an equilibrium initial shape is used, the amount of liquid transferred is still higher when the electric field is present. This suggests that even with equilibrium initial shapes, applying an electric field can still enhance the amount of liquid transferred.

4

Results: Leaky dielectrics

For leaky dielectric liquids, the governing equations are Eqs. (16)-(19). Two more parameters need to be considered: the surface charge density σ and the liquid conductivity K. The surface charge can produce a tangential stress when an electric field is present. This tangential stress can be obtained by substituting Eq. (12) into the Maxwell-stress part of Eq. (7) [31] and keeping the leading-order term. In dimensionless form we have k n · σ M · t k= σχE .

(29)

The magnitude of this stress is determined by the surface charge density (σ) and the strength of the electric field (χE) along the liquid-air interface. Note that the sign of this term depends on the signs of σ and E (since χ > 0).

4.1

Role of tangential stresses

We first study how the surface charge density and liquid conductivity influence the equilibrium bridge shape. The initial shape is a perfect cylinder with an initial aspect ratio of Λ = 1. Both the top and bottom contact angles are set to 60◦ , which means that the wettability difference ∆θr = 0. The electrocapillary number CaE is set to 0 or 50 (The electroviscous number χ in Eq. (29) is replaced by CaE when the plates are stationary (Sec. 3.3).) We set the dimensionless conductivity K ∗ to 1, and the initial surface charge density σo to 0 or 0.1 (the latter corresponds to 8.9 ×10−7 C/m2 21 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

for a potential difference of 1 kV over a 1 mm distance). The value of σo we use is smaller compared to the values used by Feng [26] when studying electrospinning, but this value is enough to have a significant influence on the bridge shape. Figure 6(a) shows that when the initial surface charge density is 0.1, the bridge shape is asymmetric even though no wettability difference is present. In contrast, the bridge shape is symmetric when no initial surface charge is present, and is very close in shape to what is observed for the corresponding perfect-dielectric case.

Figure 6: (a) Comparison of equilibrium bridge shapes between the cases where σo = 0 (solid line) and 0.1 (dashed line). The wettability difference ∆θr = 0◦ with both the top and bottom contact angles set to 60◦ . (b) Equilibrium bridge shape for the case where σ = 0.1 and both contact lines are fixed at 1. Values of other parameters are CaE = 50, K ∗ = 1, and β = 1.74. To remove the effect of contact-line motion and understand the influence of the surface charge density, we perform additional calculations in which the contact lines are fixed. Figure 6(b) shows the equilibrium bridge shape with fixed contact lines when σo = 0.1. When the bridge reaches equilibrium, the radius of the bridge near the bottom plate increases about 2%. As in Fig. 6(a), this indicates that the presence of surface charge tends to push liquid to the bottom plate. We now discuss the reason why liquid is pushed from top plate to the bottom plate when σo is sufficiently large. The key is the sign of the electrostatic contribution to the tangential stress, given by Eq. (29). When this quantity is negative, the tangential electrostatic stress points in the negative-z direction. In our calculations, 22 ACS Paragon Plus Environment

Page 22 of 42

Page 23 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

the electrocapillary number CaE and the surface charge density are positive. Because the top plate has a positive electrostatic potential and E = −∇φ, the direction of electric field points toward the bottom plate. Therefore, the direction of the tangential electrostatic stress is from the top to the bottom plate. If we want the tangential electrostatic stress to point the top plate, we would either need the initial surface charge density to be negative or we could reverse the direction of the electric field by using a positive electrostatic potential at the bottom plate while keeping the top plate grounded. This suggests that the direction of the tangential electrostatic stress can be manipulated to enhance liquid transfer to the top surface. Figure 6(a) also implies that conductivity alone does not induce enough surface charge to significantly influence bridge shape. Additional calculations we have performed indicate that even if the conductivity is increased by an order of magnitude, the bridge shape changes very little when σo = 0. In contrast, if the initial surface charge density is taken to be 0.1 and the conductivity is lowered to 0, the bridge shape closely resembles that obtained when K ∗ = 1 (Fig. 6(a)), although the asymmetry is not as pronounced. This indicates that the initial surface charge density plays a dominant role in determining bridge shape.

4.2

Liquid transfer

We now study liquid transfer in leaky dielectrics. The initial bridge is again a cylinder with an initial aspect ratio Λ = 1, the top contact angle is 60◦ while the bottom contact angle varies, and the electroviscous number χ is set to 0 or 50. From Fig. 6, we see that liquid is pushed to the bottom surface when σo = 0.1. Such an initial shape may decrease the transfer ratio relative to the case where electrostatic effects are absent. To increase the amount of liquid transferred to the top surface, the tangential electrostatic stress should point toward that surface. Therefore, here we set the initial surface charge density to σ0 = −0.1 which is opposite to the sign used in the previous section. Figure 7 shows the transfer ratio with and without the presence of an electric field at Ca = 0.1 and 1. Also plotted are the transfer-ratio values from Fig. 3 for the case 23 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7: Relationship between transfer ratio and wettability difference when χ = 0 and χ = 50 at (a) Ca = 0.1 and (b) Ca = 1. Here θtop is 60◦ and θbottom varies. Values of other parameters are K ∗ = 1, σ = -0.1, and β = 1.74. of perfect dielectrics. For leaky dielectrics, the transfer ratio increases significantly even when the bottom plate is more wettable than the top plate (∆θr < 0). When the electric field is not present and ∆θr is negative, the transfer ratio is less than 50% for both capillary numbers. However, for leaky dielectrics, the tangential electrostatic stress significantly enhances the amount of liquid transferred. Since the electric field is pointing in the negative-z direction and the sign of the surface charge is negative, the tangential electrostatic stress is always positive (pointing in the positive-z direction) and drives more liquid to the top plate. Figure 8 illustrates the electrified and non-electrified bridge shapes at different times when Ca = 0.1 and shows how the tangential electrostatic stress enhances liquid transfer to the top surface. At the beginning of stretching (Fig. 8(a)), both bridges quickly evolve to have shapes that satisfy the contact-angle boundary conditions. The bottom contact line of the electrified liquid bridge retreats faster than that of the non-electrified liquid bridge because the tangential electrostatic stress drives liquid to the top plate. At t = 0.71 (Fig. 8(b)), the bottom contact-line position of the electrified liquid bridge is only half of its initial value, and the minimum bridge radius is very close to the bottom plate. As a result, most of the liquid transfers to the top plate when the electrified bridge breaks at t = 1.1 (Fig. 8(c)). In contrast, the nonelectrified bridge breaks at t = 2.1. Thus, not only is the transfer ratio enhanced 24 ACS Paragon Plus Environment

Page 24 of 42

Page 25 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 8: Comparison of electrified and non-electrified bridge shapes at different times for the case where ∆θr = 10◦ (θbottom = 70◦ , θtop = 60◦ ) at Ca = 0.1. Values of other parameters are σo = -0.1, K ∗ = 1, and β = 1.74. The breakup time for the electrified bridge is 1.1 and the breakup time for the non-electrified bridge is 2.1.

25 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

in leaky dielectric materials, the breakup time is also reduced. The breakup time for the corresponding perfect dielectric case is 2.4, which is larger than that for the non-electrified bridge due to the electrostatic stabilization (Sec. 3.2). We now summarize the results of some additional calculations we have performed. First, if we pin the top contact line and let the bottom contact line slip (similar to Fig. S1), nearly all of the liquid is transferred to the top plate (similar to Fig. 7) due to the tangential electrostatic stress. Second, similar behavior to that shown here is observed if an equilibrium bridge is used as the initial condition rather than a perfect cylinder. Third, as in the static case, the initial surface charge density plays a more dominant role than conductivity. If the calculations are repeated for the same value of σo but with K ∗ = 0, the behavior of the transfer ratio is similar to that for the case of non-zero conductivity (with the transfer ratio being slightly smaller). If the calculations are repeated with non-zero conductivity but zero initial surface charge density, the behavior of the transfer ratio is similar to that for perfect dielectrics due to the small amount of induced charge. Thus for leaky dielectrics, electrostatic effects are strongest when both the initial surface charge density and conductivity are non-zero, and weakest when the initial surface charge density is zero. This is consistent with Eq. (29), which shows that the tangential electrostatic stress is directly proportional to the surface charge density.

5

Experiments

Although the primary purpose of this paper is to use slender-jet models to advance fundamental understanding of electrostatic assist, we have also performed some simple flow visualization experiments to further examine how electrostatic forces can influence liquid transfer.

26 ACS Paragon Plus Environment

Page 26 of 42

Page 27 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

5.1

Experimental setup

The experimental apparatus is shown in Fig. S3. This apparatus consists of two parallel plates made of polished aluminum (12.6 mm diameter; Ted Pella, Inc.) in an enclosed container. To provide a wettability contrast, for some experiments a polystyrene (PS)-covered polished aluminum bottom plate was used. The PS covered plate was fabricated by spin coating a 1 wt% solution of PS (MW 12000; Polysciences Inc.) in toluene at 2000 rpm for 60 s using a Headway Research Inc. PWM32 spinner system. The plates are separated using a motor (Parker Compumotor), and a voltage difference can be applied (Spellman SL-130 Series High Voltage Power Supply). An 88 wt% glycerol-distilled water solution is used in the experiments. Its viscosity is 119.6 ± 1.0 cP (Brookfield DV-II viscometer), surface tension is 66.5 ± 0.2 mN/m (Kruss Digital Tensiometer K10ST), relative permittivity is 45.9 ± 0.4 (Scientifica 870 Dielectric Constant Meter), and conductivity is 0.4 ± 0.1 µS/cm (Thermo Scientific Orion Conductivity Benchtop Meter). The receding contact angles of the solution on the polished aluminum specimen and PS-coated surfaces were measured to be 13◦ ±2◦ and 58◦ ± 2◦ , respectively (Kruss DSA30S). A high-speed camera (Photron FASTCAM-ultima APX with an NEC Navitar Zoom 6000 lens system) is used to visualize bridge stretching with the help of a light source on the other side of the apparatus. The experimental results are analyzed using the ImageJ software. We use Pappus’ theorem to calculate the volumes of the initial bridge and of the droplets that rest on the plates after stretching is complete [43]. Pappus’ theorem generates the volume of a liquid droplet by revolution of the droplet’s cross-sectional area in the r-z plane. This method assumes that the droplet is axisymmetric. Even though the droplets in the experiments will not be perfectly axisymmetric, the calculated total droplet volume before and after bridge stretching is typically within 5% of the expected volume of 2 µL.

27 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

5.2

Results: Two identical surfaces

As a benchmark, we first present results for the case where both plates are identical. In the absence of electrostatic forces, we would expect a transfer ratio of 50% (assuming that gravitational effects are negligible) due to the symmetry of the problem. In the presence of electrostatic forces, we would expect a transfer ratio of 50% if the liquid behaves like a perfect dielectric (Sec. 3). However, if the initial surface charge density plays a significant role, this would break the problem symmetry and we would expect a much different value of the transfer ratio (Sec. 4). We performed experiments using the polished aluminum plates. A 2 ± 0.06 µL drop of the glycerol-water solution was placed on the bottom plate. The top plate was lowered to come into contact with the droplet until the distance between two surfaces was about 0.45 mm, which enabled us to obtain initial contact radii that were approximately equal on both plates. We then increased the distance between the two plates to approximately 1.1 mm to prevent electrostatic discharge upon application of an electrostatic potential difference. After this step, an electrostatic potential difference of 1 kV (CaE ≈ 0.1) was applied in one set of experiments until the bridge reached an equilibrium (static) shape (This value of the potential difference is large enough to cause a change in bridge shape but low enough to avoid discharge.) Then, the bridge was stretched until it broke. Each experiment (with and without the electric field) was repeated at least three times. A stretching speed of approximately 6.25 mm/s was used. Scaling length with the characteristic bridge radius, which is ∼ 1 mm, yields Ca ∼ 0.01, Re ∼ 0.06, and Bo ∼ 0.1, respectively. These values imply that surface-tension forces are considerably stronger than viscous, inertial, and gravitational forces. The electroviscous number is χ ≈ 10, and the value of the electrocapillary number is CaE = χCa ≈ 0.1. The value of the latter parameter implies that although electrostatic forces are weaker than surface-tension forces, they could still influence bridge behavior. The charge relaxation time o /K is ∼10 µs, which is much smaller than the time scale for the completion of stretching, ∼ 0.1 s. This implies that conductivity will likely play a minor role. 28 ACS Paragon Plus Environment

Page 28 of 42

Page 29 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure S4 shows visualizations of static liquid bridge shapes. In the absence of an electrostatic potential difference, the narrowest bridge diameter is 1.04 mm (Fig. S4(a)). In the presence of an electrostatic potential difference, the narrowest bridge diameter is larger, 1.15 mm (Fig. S4(b)). These experimental results are qualitatively consistent with the 1D model predictions (Fig. S2). Table 4: Mean transfer ratios and bridge breakup lengths when both surfaces are identical Transfer ratio

Breakup length

φ=0V

49.7% ± 1%

1.66 ± 0.05 mm

φ = 1 kV

50.2% ± 1%

2.63 ± 0.03 mm

After characterizing static bridge shapes, experiments involving bridge stretching were performed, and the results are summarized in Table 4. When the electrostatic potential difference is absent, the mean transfer ratio is 49.7% ± 1%. When the eletrostatic potential difference is present, the mean transfer ratio is 50.2% ± 1%. Based on these results, we conclude that there is no significant difference in the transfer ratio when the plates are identical and an electric field is applied. These results suggest that the influence of surface charge (and the associated tangential stresses) are negligible in these experiments. We note that in these experiments, the contact lines appear to be pinned, possibly due to the presence of surface heterogeneities. Table 4 also shows that the bridge breakup length is larger when the electric field is present (2.63 ± 0.03 mm vs. 1.66 ± 0.05 mm). Since electrostatic forces tend to stabilize the bridge, the bridge can be stretched longer without breaking compared to the case where the electric field is absent. These results are qualitatively consistent with the predictions of the perfect dielectric model (Fig. 3(a)).

5.3

Results: Two different surfaces

We now consider the case where the two surfaces are different. Polished aluminum is used for the top plate, and PS-coated aluminum is used for the bottom plate. We 29 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

found that by using the method described in Sec. 5.2, we had difficulty obtaining initial bridge shapes that were consistent for each run, presumably due to surface heterogeneities. Instead, we used a slightly different method for creating the bridges. A 2 ± 0.06 µL drop of the glycerol-water solution was placed on the bottom plate. The top plate was lowered until the distance between the two plates was approximately 1 mm (This distance was controlled by plastic shims (Artus Corp.)) At this distance, the top plate is not yet in contact with the droplet. An electrostatic potential difference of about 1 kV was then applied, causing the droplet to stretch and come into contact with the top plate, forming a liquid bridge. In some experiments, the electrostatic potential was then turned off until the bridge reached an equilibrium shape and then the bridge was stretched. In other experiments, the electrostatic potential difference was increased to 1.5 kV (CaE ≈ 0.27), an equilibrium bridge shape was reached, and then the bridge was stretched. Each set of experiments was performed at least three times. We used the same stretching speed as in Sec. 3.2, so surface-tension forces are again expected to play a key role. Figure 9 shows images of bridge shapes before stretching and right after breakup. Table 5 summarizes the results. In the experiments without the electric field, both top and bottom contact lines are pinned during stretching, which we suspect occurs due to the presence of surface heterogeneities. In the experiments with the electric field, the top contact line remains pinned but the bottom contact line retreats and then expands during stretching. The transfer ratio without the electric field is 34% ± 2%, and the transfer ratio with the electric field is 42% ± 2%. These results show that the transfer ratio increases when the electric field is present. Furthermore, the average breakup length of the non-electrified bridge is 2.06 ± 0.01 mm (Fig. 9(b)), but the average breakup length of the electrified bridge is longer, 3.03 ± 0.08 mm (Fig. 9(d)). We now discuss how the electric field influences bridge deformation and thus the transfer ratio. Before the bridge is stretched, the shapes of both liquid bridges are slightly different. The mean top and bottom contact radii of both bridges are 0.80 ± 0.03 mm and 1.10 mm ± 0.02 mm, respectively. The mean narrowest bridge 30 ACS Paragon Plus Environment

Page 30 of 42

Page 31 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 9: Visualizations of the liquid bridge (a) before stretching and (b) after liquid bridge breaks without application of an electrostatic potential difference (χ = 0). Visualizations of the liquid bridge (c) before stretching and (d) after application of an electrostatic potential difference of 1.5 kV (χ ≈ 27).

Table 5: Contact line motion, transfer ratio and the breakup length in the experiments with two different surfaces Top contact line

Bottom contact line

Transfer ratio

Breakup length

φ=0V

Pinned

Pinned

34% ± 2%

2.06± 0.01 mm

φ = 1.5 kV

Pinned

Free

42% ± 2%

3.03 ± 0.08 mm

31 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

diameter of the electrified bridge is 1.54 ± 0.01 mm, which is larger than that of the non-electrified bridge (1.46 ± 0.03 mm) (Figs. 9(a) and (c)). During stretching, both contact lines of the non-electrified bridge are pinned. Therefore, the contact radii at the upper and lower plates have the same values after bridge breakup. For the electrified bridge, the top contact line does not slip, but the bottom contact line retreats at the beginning of stretching. The bottom contact line then expands when the bridge is about to break. The final bottom contact radius is slightly smaller (∼ 0.03 mm) than its initial value. The top contact lines for both cases are pinned. This pinning may be due to the receding contact angle of the polished aluminum surface being so small (13◦ ) that the bridge breaks before the top contact angle reaches the receding value. When the electric field is not present, the bottom contact line is pinned, which may also be due to the small contact angle. However, the bottom contact line retreats for the electrified bridge. When the electric field is present, electrostatic forces stabilize the bridge by increasing the narrowest bridge radius. Liquid must flow from the region near both contact lines to the middle of the bridge, so both the top and bottom contact angles decrease before stretching. As a consequence, the bottom contact angle reaches its receding value during stretching and the bottom contact line retreats, causing the transfer ratio to increase.

5.4

Comparison to model predictions

The experimental results suggest that the presence of the electric field can significantly influence contact-line motion. Thus, in comparing our model predictions to the experimental results, we examine all combinations of boundary conditions to advance fundamental understanding. These include the cases where (i) both the top and bottom contact lines are free to move, (ii) both the top and bottom contact lines are pinned, (iii) the top contact line is pinned and the bottom contact line is free to move, and (iv) the top contact line is free to move and the bottom contact line is pinned. 32 ACS Paragon Plus Environment

Page 32 of 42

Page 33 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 10: (a) Initial bridge shapes and (b) final bridge shapes with (χ = 27; dashed line) and without (χ = 0; solid line) the electric field. Both top contact radii are fixed at 0.8 mm. The bottom contact radius for the case where χ = 0 is fixed at 1 mm, while the bottom contact radius for the case where χ = 27 is free to move and the contact angle is 58◦ . Here, β = 44.9. Figure 10(a) shows the initial bridge shapes used in our calculations. The initial shapes represent static bridge shapes and are generated using the procedure described in Sec. 3.3. The radii at the top and bottom plates have been chosen to match those in the experiments. As can be seen in Fig. 10(a), the contact angles at both the top and bottom plates of the electrified bridge are smaller than those of the non-electrified bridge. Also, the narrowest bridge radius of the electrified bridge is (0.74 mm) is slightly greater than that of the non-electrified bridge (0.73 mm). These values are comparable to what is observed in the experiments (0.77 ± 0.01 mm for the electrified bridge and 0.73 ± 0.01 mm for the non-electrified bridge); the discrepancy may be due to the lack of perfect axisymmetry in the experiments. Table 6 shows the calculated transfer ratios and breakup lengths for different boundary conditions. The entries in boldface correspond to the conditions observed in the experiments. We first compare these predictions with the experimental results, then discuss the influence of different boundary conditions. In the experiments without the electric field, both contact lines are pinned during stretching. The transfer ratio and breakup length predicted by the 1D model with these boundary conditions are 34.72% and 2.03 mm, respectively. These results agree well with the experimental results shown in Table 5. In the experiments with the 33 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 42

Table 6: Predicted transfer ratio and breakup length for different boundary conditions. The entries in boldface correspond to the conditions observed in the experiments. Electric potential

Contact line boundary conditions Top contact line

Bottom contact line

Transfer ratio

Breakup length

Pinned

Pinned

34.72%

2.03 mm

Pinned

Free

43.76%

2.16 mm

Free

Pinned

62.92%

1.57 mm

Free

Free

95.85%

1.35 mm

Pinned

Pinned

34.72%

2.03 mm

Pinned

Free

45.45%

2.25 mm

Free

Pinned

63.43%

1.57 mm

Free

Free

95.93%

1.35 mm

φ=0V

φ = 1.5 kV

electric field, the top contact line is pinned but the bottom contact line is free to move. Here, the predicted transfer ratio is 45%, which is close the experimental value of 42 ± 2%. In the calculations, the bottom contact line first retreats and then expands when the liquid bridge is close to pinchoff. At the end of stretching, the bottom contact radius of the electrified liquid bridge decreases from 1 mm to 0.97 mm, which agrees well with the experimental result (0.97 mm). Figure 10(b) shows the predicted bridge shapes after breakup. Note that the pinch-off points of both bridges are nearly the same, but the final length of the electrified bridge (2.25 mm) is larger than that of the non-electrified bridge (2.03 mm). Because the electrostatic forces decrease the pressure difference between the narrowest region of the liquid bridge and the bottom contact line as discussed in Sec. 3.2, the pinch-off point of the electrified bridge is closer to the bottom surface (Fig. 10(b)). Therefore, although the bottom contact line position decreases just 0.03 mm, the transfer ratio can increase by approximately 8%. However, the predicted breakup length of the electrified bridge (2.25 mm) is smaller than the experimental value (3.03 mm). Since the transfer ratio is set by 34 ACS Paragon Plus Environment

Page 35 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

the overall bridge shape rather than the breakup length, the transfer ratio predicted by the 1D model still agrees well with the experiments. The discrepancy in the breakup length between experiment and theory may be due to neglecting the effect of the normal component of the electric field in the 1D model. Capturing this effect would require development of a full 2D model, a task we leave for future work. Table 6 illustrates that the contact-line boundary conditions significantly influence the transfer ratio. Whether or not electrostatic effects are present, allowing one of the contact lines to move increases the transfer ratio. Allowing the bottom contact line to move has a larger influence, presumably due to the fact the narrowest diameter of the bridge is located near the bottom plate. The largest transfer ratio (and smallest breakup length) occurs when both contact lines are free to move. Notably, even when the electric field is absent, allowing the bottom contact line to move yields a transfer ratio close to the experimental value. This indicates that the primary influence of electrostatics is through its influence on the boundary conditions. Indeed, for the same boundary conditions, application of the electric field leads to an increase in the transfer ratio by only a modest amount. This can be rationalized by noting that the ratio of electrostatic to surface-tension forces, which is given by χCa, is equal to 0.27. Unfortunately, we were unable to further increase the strength of the electric field in the experiments due to electrostatic discharge. If we attempt to increase the capillary number by raising the liquid viscosity or stretching speed, the contact lines on both surfaces tend to pin since the stronger viscous forces oppose the capillary pressure gradients that drive contact-line slip. Nevertheless, application of the electric field still has a significant influence on the transfer ratio since it leads to depinning of the contact line on the bottom plate. As discussed above, application of the electric field modifies the interface shape and causes the contact angle to drop below the receding value, thereby depinning the contact line.

35 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

6

Conclusions

In this work we have used slender-jet models to advance fundamental understanding of how electric fields can be used to improve liquid transfer. Application of an electric field lowers the pressure inside the liquid bridge. This causes the pressure difference between the bridge center and the more wettable plate to become larger relative to the pressure difference between the bridge center and the less wettable plate. As a result, more liquid is driven toward the more wettable plate. For leaky dielectrics, the electrostatic contribution to the tangential stress also plays a key role. If the sign of the initial surface charge density and the direction of the electric field are chosen appropriately, nearly all the liquid can be driven to the more wettable plate. More generally, our results suggest that the surface charge and electric field can be used to augment or oppose flows driven by wettability differences. One of major challenges in high-speed printing is obtaining a transfer ratio greater then 50%. Our results indicate that application of electric fields can help overcome this challenge, and our work thus advances fundamental understanding of electrostatic assist (ESA), a method used in inudstrial printing processes to improve liquid transfer. Results from our flow visualization experiments reveal that electric fields can also significantly influence contact-line motion. Surface heterogeneities, which are not present in our models, are likely important in the experiments and can lead to contact-line pinning. Electric fields can deform the liquid-air interface and cause the contact lines to unpin. By accounting for the different boundary conditions observed in the experiments, our models yield predictions consistent with the experimental observations. Gravure printing, an industrial process to which ESA is applied, involves liquid transfer from a cavity to a flat surface. Describing liquid transfer in this configuration is beyond the reach of the slender-jet approach and requires full 2D (and even 3D [44]) models. Such models, along with more comprehensive experiments, will be necessary to more effectively apply ESA in the manufacturing of printed electronic devices.

36 ACS Paragon Plus Environment

Page 36 of 42

Page 37 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Acknowledgments This work was supported through the Industrial Partnership for Research in Interfacial and Materials Engineering of the University of Minnesota. We thank Wieslaw Suszynski for his assistance with the experiments.

Supporting Information Figures S1, S2, S3, S4

37 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

References [1] M. Pudas, J. Hagberg, and S. Leppavuori, Printing parameters and ink components affecting ultra-fine-line gravure-offset printing for electronics applications, J. Eur. Ceram. Soc., 24:2943–2950, 2004. [2] B. Kang, W. H. Lee, and K. Cho Recent Advances in Organic Transistor Printing Processes, Appl. Mater. Interfaces, 5:2302–2315, 2013. [3] P. F. Moonen, I. Yakimets, and J. Huskens, Fabrication of transistors on flexible substrates: From mass-printing to high-resolution alternative lithography strategies, Adv. Mater., 24:5526–5541, 2012. [4] H. Kang, R. Kitsomboonloha, K. Ulmer, L. Stecker, G. Grau, J. Jang, V. Subramanian, Megahertz-class printed high mobility organic thin-film transistors and inverters on plastic using attoliter-scale high-speed gravure-printed sub-5 µm gate electrodes, Org. Electron., 15:3639–3647, 2014. [5] F. C. Krebs, Fabrication and processing of polymer solar cells: A review of printing and coating techniques, Sol. Energ. Mat. Sol. Cells, 93:394412, 2009. [6] M. Jung, J. Kim, J. Noh, N. Lim, C. Lim, G. Lee, J. Kim, H. Kang, K. Jung, A. D. Leonard, J. M. Tour, and G. Cho, All-Printed and roll-to-roll-printable 13.56-MHz-Operated 1-bit RF tag on plastic foils, IEEE Trans. Electron Devices, 57:571–580, 2010. [7] D. Tobjork and R. Osterbacka, Paper Electronics, Adv. Mater., 23:1935–1961, 2011. [8] S. Khan, L. Lorenzelli, and R. S. Dahiya, Technologies for printing sensors and electronics over large flexible substrates: A review, IEEE Sensors J., 15:3164– 3185, 2015. [9] S. Kumar, Liquid transfer in printing processes: Liquid bridges with moving contact lines, Annu. Rev. Fluid Mech., 47:67–94, 2015. 38 ACS Paragon Plus Environment

Page 38 of 42

Page 39 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

[10] D. Sung, A. F. Vornbrock, and V. Subramanian, Scaling and optimization of gravure-printed silver nanoparticle lines for printed electronics, IEEE Trans. Compon. Packag. Technol., 33:105–114, 2010. [11] A. V. Joshi, C. Dettke, and J. Steingraeber, Investigation on electrostatic assist and gravure process parameters on solid mottle reduction for shrink films, J. Coat. Technol. Res., 13:375–383, 2016. [12] H. F. George, Electrostatically assisted ink transfer in gravure printing, ACS Symp. Ser., 200:359–370, 1982. [13] K. Morris, Electrostatically assisted gravure, Print. Technol., 12:180, 1968. [14] A. Ramkrishnan and S. Kumar, Electrohydrodynamic deformation of thin liquid films near surfaces with topography, Phys. Fluids 26:122110, 2014. [15] M. J. Vogel and P. H. Steen, Capillarity-based switchable adhesion, Proc. Natl. Acad. Sci., 8:3377–3381, 2010. [16] S. Gaudet, G. H. McKinley, and H. A. Stone, Extensional deformation of Newtonian liquid bridges, Phys. Fluids, 8:2568–2579, 1996. [17] T. M. Squires and S. R. Quake, Microfluidics: Fluid physics at the nanoliter scale, Rev. Mod. Phys., 77:977–1026, 2005. [18] H. Chen, T. Tang, and A. Amirfazli, Liquid transfer mechanism between two surfaces and the role of contact angles, Soft Matter, 10:2503–2507, 2014. [19] H. Chen, T. Tang, and A. Amirfazli, Effects of surface wettability on fast liquid transfer, Phys. Fluids, 27:112102, 2015. [20] H. Chen, T. Tang, and A. Amirfazli, Fast liquid transfer between surfaces: breakup of stretched liquid bridges, Langmuir, 31:11470-11476, 2015. [21] S. Dodds, M. S. Carvalho, and S. Kumar, Stretching and slipping of liquid bridges near plates and cavities, Phys. Fluids, 21:092103, 2009.

39 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[22] C.-H. Huang, M. S. Carvalho, and S. Kumar Stretching liquid bridge with moving contact lines: Comparison of liquid-transfer predictions and experiments, Soft Matter, 12:7457–7469, 2016. [23] S. Dodds, M. S. Carvalho, and S. Kumar, Stretching liquid bridges with moving contact lines: The role of inertia, Phys. Fluids, 23:092101, 2011. [24] X. Zhang, R. S. Padgett, and O. A. Basaran, Nonlinear deformation and breakup of stretching liquid bridges, J. Fluid Mech., 329:207–245, 1996. [25] B. Qian and K. S. Breuer, The motion, stability and breakup of a stretching liquid bridge with a receding contact line, J. Fluid Mech., 666:554–572, 2011. [26] J. J. Feng, The stretching of an electrified non-Newtonian jet: A model for electrospinning, Phys. Fluids, 14:3912–3926, 2002. [27] M. M. Hohman, M. Shin, G. Rutledge, and M. P. Brenner Electrospinning and electrically forced jets. I. Stability theory, Phys. Fluids, 13:2201–2220, 2001. [28] J. J. Feng, Stretching of a straight electrically charged viscoelastic jet, J. NonNewtonian Fluid Mech., 116:55–70, 2003. [29] G. Taylor, Studies in electrohydrodynamics. I. The circulation produced in a drop by electrical field, Proc. R. Soc. A, 291:159–166, 1966. [30] J. R. Melcher and G. I. Taylor, Electrohydrodynamics: A review of the role of interfacial shear stresses, Annu. Rev. Fluid Mech., 1:111–146, 1969. [31] D. A. Saville, Electrohydrodynamics: The Taylor-Melcher leaky dielectric model, Annu. Rev. Fluid Mech., 29:27–64, 1997. [32] A. Ramkrishnan and S. Kumar, Electrohydrodynamic effects in the leveling of coatings, Chem. Eng. Sci., 101:785–799, 2013. [33] J. Eggers, Nonlinear dynamics and breakup of free-surface flows, Rev. Mod. Phys., 69:865-929, 1997.

40 ACS Paragon Plus Environment

Page 40 of 42

Page 41 of 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

[34] J. Eggers and T. F. Dupont, Drop formation in a one-dimensional approximation of the Navier-Stokes equation, J. Fluid Mech., 262:205–221, 1994. [35] O. E. Yildirim and O. A. Basaran, Deformation and breakup of stretching bridges of Newtonian and sheer-thinning liquids: Comparison of one- and twodimensional models, Chem. Eng. Sci., 56:211–233, 2001. [36] J. Eggers and E. Villermaux, Physics of liquid jets, Rep. Prog. Phys., 71:036601, 2008. [37] H. A. Stone, J. R. Lister, and M. P. Brenner Drops with conical ends in electric and magnetic fields, Proc. R. Soc. Lond. A, 455:329–347, 1999. [38] S. Sankaran and D. A. Saville Experiments on the stability of a liquid bridge in an axial electric field, Phys. Fluids, 5:1081:1083, 1993 [39] A. Ramos, H. Gonzalez, and A. Castellanos, Experiments on dielectric liquid bridges subjected to axial electric fields, Phys. Fluids, 6:3206–3208, 1994. [40] C. L. Burcham and D. A. Saville The electrohydrodynamic stability of a liquid bridge: Microgravity experiments on a bridge suspended in a dielectric gas, J. Fluid Mech., 405:37–56, 2000. [41] C. L. Burcham and D. A. Saville Electrohydrodynamic stability: Taylor-Melcher theory for a liquid bridge suspended in a dielectric gas, J. Fluid Mech., 452:163– 187, 2002 [42] C. Gupta, G. A. Mensing, M. A. Shannon, and P. J. A. Kenis, Double transfer printing of small volumes of liquids, Langmuir, 23:2906-2914, 2007. [43] R. McLenaghan and S. Levy, CRC Standard Mathematical Tables and Formulae, CRC Press, 1995. [44] S. Dodds, M. S. Carvalho, and S. Kumar, The dynamics of three-dimensional liquid bridges with pinned and moving contact lines, J. Fluid Mech., 707:521-540, 2012.

41 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For Table of Contents Only

42 ACS Paragon Plus Environment

Page 42 of 42